Statistics Colloquium : Dr. Edward Kennedy
CMU
Location
Online
Date & Time
March 4, 2022, 11:00 am – 12:00 pm
Description
Title: Nonparametric estimation of heterogeneous causal effects
Abstract: Heterogeneous effect estimation plays a crucial role in causal inference, with applications across medicine and social science. Many methods for estimating conditional average treatment effects (CATEs) have been proposed in recent years, but there are important theoretical gaps in understanding if and when such methods are optimal. This is especially true when the CATE has nontrivial structure (e.g., smoothness or sparsity). This talk surveys work across two recent papers in this context. First, we study a two-stage doubly robust CATE estimator and give a generic model-free error bound, which, despite its generality, yields sharper results than those in the current literature. The second contribution is aimed at understanding the fundamental statistical limits of CATE estimation. To that end, we resolve this long-standing problem by deriving a minimax lower bound, with matching upper bound based on higher order influence functions. We illustrate with examples from infectious disease epidemiology and voter turnout experiments in political science.
Abstract: Heterogeneous effect estimation plays a crucial role in causal inference, with applications across medicine and social science. Many methods for estimating conditional average treatment effects (CATEs) have been proposed in recent years, but there are important theoretical gaps in understanding if and when such methods are optimal. This is especially true when the CATE has nontrivial structure (e.g., smoothness or sparsity). This talk surveys work across two recent papers in this context. First, we study a two-stage doubly robust CATE estimator and give a generic model-free error bound, which, despite its generality, yields sharper results than those in the current literature. The second contribution is aimed at understanding the fundamental statistical limits of CATE estimation. To that end, we resolve this long-standing problem by deriving a minimax lower bound, with matching upper bound based on higher order influence functions. We illustrate with examples from infectious disease epidemiology and voter turnout experiments in political science.